Huynh Hieu Travel là một trang blog du lịch. Chúng tôi cung cấp cho bạn những thông tin, kinh nghiệm và review về những địa điểm du lịch thú vị, trải nghiệm.
Fun interactive science app ideas | Part 5
Built a periodic table app that visualizes atomic and molecular structures
UI Design
GPT Images 2
Code
Gemini 3.1 Pro
More demos ↓
anthropic acaba de cambiar cómo se diseña frontend con IA.
se llama Frontend Design, es su skill oficial para Claude Code.
y ya tiene 136.000 estrellas en GitHub.
el problema que resuelve es simple:
cada vez que le pedías a la IA una web, te salía lo mismo.
inter como fuente.
gradiente morado.
layout genérico de startup.
el "AI slop" de siempre.
Frontend Design lo corta de raíz.
obliga a Claude a comprometerse con una dirección estética antes de tocar una sola línea de código.
y el resultado cambia completamente.
puedes pedirle:
→ brutalista
→ editorial
→ retro-futurista
→ lujoso
→ maximalista
y genera HTML, CSS, JS, React o Vue listos para producción.
cómo instalarlo:
1. npx skills add https://t.co/h163bYeZeJ --skill frontend-design
2. describe la web o componente que quieres.
3. elige el estilo.
4. Claude monta tipografía, color, motion y composición coherente.
una sola instalación. disponible en todas tus sesiones.
funciona en Claude Code, Cursor, Codex, Gemini CLI, OpenCode y 15+ harnesses.
100% open source. mantenida por Anthropic.
418.000 instalaciones en pocas semanas dicen más que cualquier review.
repo 👇
Fun interactive science app ideas | Part 3
Played around with generating 3D biological structures and made an app to explore them interactively
UI Design
GPT Images 2
Code
Gemini 3.1 Pro
More demos ↓
@MrBuLoong@lsminhphuong Xe hơi chặn đường, tàu điện đường ngắn dừng để đợi xử lý.
Nhưng chệnh time nên tàu lớn chạy đến đụng.
Đọc kỹ bài báo nước ngoài sẽ hiểu nè bạn, đưa báo Việt ra sao đủ ngữ cảnh được.
how to use autoreason for on-page SEO / GEO
you have a blog post sitting at position 6 for a 2,400/mo keyword, 1.4% CTR, pulling 180 visits a month that converts at 0.8%. you want all three numbers to improve without rewriting the page from scratch. most marketers would prompt an LLM with "rewrite this for SEO" and publish whatever comes back.
the problem with asking AI to rewrite its own work is it never says "this is already good." it invents missing sections, drifts from the original angle with every pass, and strips out the lines that were ranking.
autoreason fixes that with adversarial isolation. every role in the loop is a fresh agent that cant see what the others wrote, so the output doesnt collapse into the same generic advice showing up in the 10 posts already outranking you.
> incumbent is your current page, the hero, the subhead, the H2 structure, the section order
> strawman is a fresh agent that sees only the page + target keyword + top 3 SERP competitors + what your ICP asked about in sales calls, it writes critiques like "H2 on line 142 is a restatement not an entity, the example in section 3 doesnt match the ICP vocabulary from the transcripts." no rewriting, just finding problems
> author B is a fresh agent that sees the task + the original + the critique and writes a revised version
> synthesizer is a fresh agent that sees original + B in random order and merges the strongest parts of both
> judge panel of 3 fresh agents does a blind ranked-choice vote on A, B, and AB using Borda count, convergence at streak=2 so it doesnt loop forever
for on-page work the knowledge layer is what keeps this grounded in your business instead of generic SEO advice:
> Google Search Console so the agent knows which queries youre getting impressions on but losing CTR, which keywords youre ranking for without optimizing for, and which pages are declining month over month
> Ahrefs or Moz data on the top 3 ranking competitors, their internal link patterns, the anchor text they use, the entities they cover that you dont
> GA4 conversion data so the judges can score variants on "would the ICP reading this still convert" not just "does it look better"
> sales call transcripts and support tickets so the agent writes in the language your customers use, not the language your marketing team assumes they use
> your own top-performing pages so the synthesizer can merge in patterns that work for your domain instead of copying whatever is on the SERP right now
this works the same way for service pages, free tool pages, and programmatic SEO templates:
> service pages: loop the "we do X for Y" page, judges score on "would a prospect forward this internally" using won-deal language from your CRM as ground truth
> free tool pages (the ones youre vibe coding with Claude): loop the landing copy around the tool, judges score against the queries driving impressions in GSC and how competing tool pages structure above the fold
> programmatic SEO templates: loop the template, not each page, judges score a sample of 20 filled pages for entity coverage, uniqueness, and conversion fit. fix the template once, 10,000 pages improve
> AI overviews and LLM citations (GEO): judges are actual LLMs prompted with the target query, checking whether your page gets cited. if it doesnt, the critic flags buried claims and weak entity definitions and the loop rewrites until it does
you still push the winner live and watch it in Search Console for 14 days before calling it. autoreason just narrows the search space so you´re shipping strong candidates instead of whatever ChatGPT cleaned up this morning.
results feed back every run so the next one is sharper
🚨 son dakika chat-gpt codex ile aynı anda hem oyunu kodluyor, hem oynuyor, hem de gerçek zamanlı olarak oyunda değişiklik yapıp; yapılanları gözleyebiliyorsunuz.
Im spechlesss!
Since the last post did really well, I’m dropping the exact prompt I used 👇
GPT’s Images2 × Seedance 2.0 workflow is powerful when used right
1️⃣ Combine your idea with the ✅ prompt
2️⃣ Let GPT generate the storyboard image
3️⃣ Use that image as a reference in Seedance
4️⃣ Paste the refined prompt back into Seedance
This is the base I used to get consistent results.
Still not perfect — motion and camera variety can feel limited — but with the right prompt structure, the quality jumps a lot.
Next step is pushing this beyond “template feel” into something more cinematic.
Prompt ⬇️
Okay I owe you guys an apology.
I've been using AI wrong this entire time. literally burning hours in figma trying to build sections when Claude Design does the same thing in minutes.
And somehow it's responsive??.
Take my money Claude, all of it.
I'll build out more sections, and share progress with you fine people. 🎯
Yeah, whole stack is getting redone.
#claude
Tú también puedes crear una landing page de 50.000$.
Sin agencia. Sin saber código. Sin presupuesto.
Te dejo por aquí un tutorial de 25 minutos para hacerlo tú mismo con Gemini + Nano Banana.